Download pdf - Electrically-tunable pore morphology in nanoporous … · Electrically-tunable pore morphology in nanoporous gold thin films Tatiana S. Dorofeeva1, and Erkin Seker1( ) Nano Res

Nano Res

1

Electrically-tunable pore morphology in nanoporous

gold thin films

Tatiana S. Dorofeeva1, and Erkin Seker1( )

Nano Res., Just Accepted Manuscript • DOI 10.1007/s12274-015-0726-x

http://www.thenanoresearch.com on January 28, 2015

© Tsinghua University Press 2015

Just Accepted

This is a “Just Accepted” manuscript, which has been examined by the peer-review process and has been

accepted for publication. A “Just Accepted” manuscript is published online shortly after its acceptance,

which is prior to technical editing and formatting and author proofing. Tsinghua University Press (TUP)

provides “Just Accepted” as an optional and free service which allows authors to make their results available

to the research community as soon as possible after acceptance. After a manuscript has been technically

edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP

article. Please note that technical editing may introduce minor changes to the manuscript text and/or

graphics which may affect the content, and all legal disclaimers that apply to the journal pertain. In no event

shall TUP be held responsible for errors or consequences arising from the use of any information contained

in these “Just Accepted” manuscripts. To cite this manuscript please use its Digital Object Identifier (DOI®),

which is identical for all formats of publication.

Nano Research

DOI 10.1007/s12274-015-0726-x

Template for Preparation of Manuscripts for Nano Research

TABLE OF CONTENTS (TOC)

Electrically-Tunable Pore Morphology in Nanoporous

Gold Thin Films

Tatiana S. Dorofeeva, Erkin Seker *

University of California – Davis, USA

Electrical current is used for tuning pore morphology of

nanoporous gold thin films at significantly lower temperatures

than previously reported via electrically-assisted mechanisms.

This technique allows for precisely controlling the extent and

location of pore coarsening and producing a wide range of

distinct morphologies on a single substrate for high-throughput

studies of structure-property relationships.

Erkin Seker, http://web.ece.ucdavis.edu/~eseker/index.html

http://web.ece.ucdavis.edu/~eseker/index.html


gold thin films

Tatiana S. Dorofeeva1, and Erkin Seker1( )

Received: day month year

Revised: day month year

Accepted: day month year

(automatically inserted by

the publisher)

© Tsinghua University Press

and Springer-Verlag Berlin

Heidelberg 2014

KEYWORDS

nanoporous gold,

tunable morphology,

Joule heating,

material screening

ABSTRACT

Nanoporous gold (np-Au) is an emerging nanostructured material that exhibits

many desirable properties, including high electrical and thermal conductivity,

high surface area-to-volume ratio, tunable pore morphology, well-established

surface-binding chemistry, and compatibility with microfabrication. These

features made np-Au a popular material for fuel cells, optical and electrical

biosensors, drug delivery vehicles, neural electrode coatings, and as a model

system for nano-scale mechanics. In each application, np-Au morphology plays

an essential role in the overall device operation. Therefore, precise control of

morphology is necessary for attaining optimal device performance.

Traditionally, thermal treatment in furnaces and on hot plates is used for

obtaining np-Au with self-similar but coarser morphologies. However, this

approach lacks the ability to create different morphologies on a single substrate

and requires high temperatures (>250 °C) that are not compatible with most

plastic substrates. Herein, we report electro-annealing as a method that for the

first time makes it possible to control the extent and location of pore coarsening

on a single substrate with one fast treatment step. In addition, the

electro-annealing entails much lower temperatures (<150 °C) than traditional

thermal treatment, putatively due to electrically-assisted phenomena that

contribute to thermally-activated surface diffusion of gold atoms responsible

for coarsening. Overall, this approach can be easily scaled up to display

multiple pore morphologies on a single chip and therefore enable

high-throughput screening of optimal nanostructures for specific applications.

1. Introduction

Nanostructured materials have attracted significant

interest from academia and industry alike, due to

intriguing fundamental questions around nanoscale

phenomena and applications enabled by properties

unique to this length scale. Optical, electrical,

magnetic, thermal, mechanical properties become

Nano Research

DOI (automatically inserted by the publisher)

Address correspondence to Erkin Seker, [email protected]

Research Article

mailto:[email protected]

| www.editorialmanager.com/nare/default.asp

2 Nano Res.

increasingly size-dependent at nanoscale [1].

Nanostructured materials have been shown to

exhibit high strength [2], unusual thermal and

electrical properties [3, 4], and enhanced surface

plasmons [5], highlighting the importance of

controlling nanostructure. An emerging

nanostructured material is nanoporous gold

(np-Au), for which related applications and

publications have rapidly grown during last decade

[6]. This popularity is due to many of its attractive

properties, including high electrical and thermal

conductivity, high surface area-to-volume ratio,

tunable pore morphology, well-established gold

binding chemistry, and compatibility with

microfabrication [6]. These properties had an

impact on applications including sensing [7, 8].

biomedical devices [9-12], energy storage [13],

photonics [14, 15], catalysis [16, 17], as well as

fundamental structure-property studies [18-21].

Nanoporous gold is typically obtained by a process

known as dealloying [22]. In this process, selective

removal of silver atoms from a silver-rich gold alloy,

accompanied by surface diffusion of gold atoms at

the metal-electrolyte interface, results in an

open-cell structure with interconnected ligaments of

tens of nanometers [22, 23]. Alloy composition [24]

and preparation approach, dealloying method

[25-27], and post-processing techniques [28, 29] all

influence pore morphology, resulting in a wide

range of possible pore and ligament sizes ranging

from tens of nanometers [25] up to several hundred

of nanometers [23]. When np-Au films are subjected

to homogenous thermal treatment (such as in

annealing furnaces), pore and ligament sizes

increase in a uniform fashion across the entire

sample due to enhanced diffusion of surface atoms

[23]. However, a controllable spatial gradient of

pore morphology is highly desirable in order to

rapidly identify optimal morphologies for desired

application [29]. For example, it is often necessary

to identify a specific range of nanostructures that

increase surface-enhanced Raman signal from a

metallic surface, improve biocompatibility, lower

limit-of-detection of a sensor, or increase reactivity

of a catalytic surface. With the uniform exposure to

thermal energy in conventional annealing

approaches, it is not possible to obtain different

pore morphologies on a single substrate. In

addition, traditional thermal treatment in furnaces

or on hot plates requires temperatures of at least

250 ˚C [23]; thereby, rendering this technique

incompatible with most plastic substrates with low

melting temperatures [30].

Here, we present a novel technique to address these

challenges by taking advantage of the electrical

conductivity of np-Au thin films in order to tune

pore morphology. Specifically, we employ

electro-annealing of micropatterned np-Au thin film

patterns with rationally-designed geometries to

obtain distinct pore morphologies on a single chip.

This technique is based on the principle that

electrical current flowing in thin film traces lead to

localized changes in current density as a function of

trace geometry. At regions with smaller

cross-section, increased current density leads to

localized Joule heating (also known as ohmic and

resistive heating) and putatively enhances

electrically-assisted coarsening mechanisms (to be

discussed in section 3.4). The interplay of thermal

and electrical mechanisms consequently produces

coarsened morphologies at temperatures much

lower (~150 ˚C) than those required for pure

thermal annealing [23], thus opening the door to the

use of substrates with lower melting temperatures.

In tandem, this method, for the first time, allows for

creating multiple and distinct pore morphologies

on a single substrate via a short (less than 2

minutes), single treatment step. Here, we describe

the electro-annealing technique and illustrate

resultant pore morphologies. Finally, a comparative

study on differences between temperature and

morphology for electro- and hot plate-annealing

approaches is presented. We expect that this novel

approach will facilitate the production of

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research

3 Nano Res.

high-throughput screening platforms on various

substrates for studying structure-property

relationships and identifying optimal morphologies

for desired applications.

2. Experimental

In order to study the effect of thin film geometry on

pore morphology evolution, we fabricated samples

with various np-Au thin film patterns on glass

coverslips: (i) simple rectangular traces, referred to

as unconstricted samples (Figure 1A); (ii) geometric

variations that act as electrical current concentrators

within the rectangular traces, referred to as

constricted samples (Figure 3A); and (iii) branched

trace structures with constrictions that act as

current dividers controlling the extent of

electro-annealing (Figure 4A). The np-Au traces

(460 nm-thick) on glass coverslips were produced

by dealloying lithographically-patterned gold-silver

thin films that were sputter-deposited, as

previously described [31]. The prepared samples

were mounted on a homemade electro-annealing

test fixture (Figure S-1 in the ESM), constant current

was injected into the trace through the clips, and the

sample temperature was measured by an infrared

(IR) thermometer positioned over the sample

(Figure S-2 in the ESM). High-magnification sample

images were obtained with a scanning electron

microscope (SEM) and the median ligament

thicknesses were extracted from each image to

quantify morphology evolution.

Chemicals and Materials: Glass coverslips (24 x 60

mm, thickness 0.13 - 0.16 mm) were purchased from

Fisher Scientific. Piranha solution, used for cleaning

the coverslips, consisted of 4:1 mixture of hydrogen

peroxide (concentration 30%) and sulfuric acid

(concentration 96%). Positive photoresist (1813) and

developer (MF-322) were bought from Shipley.

Metal targets (Cr, Au, and Ag of 99.95% purity)

were purchased from Kurt J. Lesker.

N-methylpyrrolidone (NMP) was used for lift-off

and nitric acid (concentration 70%) was used for

dealloying.

Sample preparation: The gold-silver alloy (precursor

to np-Au) patterns were created on the glass

coverslips with a combination of photolithography,

sputter-deposition, and dealloying processes [31].

Briefly, acid-cleaned glass coverslips were coated

with 1 μm-thick positive photoresist that was

exposed through a transparency mask and

developed to form the lift-off layer for subsequent

metal deposition. The gold-silver micropatterns

were obtained by sequential deposition of 120 nm

of chrome (adhesion layer), 80 nm of gold (seed

layer), and 500 nm of co-sputtered gold and silver.

Following lift-off patterning by immersion in NMP,

the samples were dealloyed in 70% nitric acid for 15

minutes at 55 ˚C to produce the np-Au films. The

samples were then rinsed and soaked in DI water

for two days to remove any residual nitric acid.

CAUTION: Both piranha and nitric acid solutions

are highly corrosive and should be handled with

care. Please see the Supplementary Information for

details of the fabrication process.

Electro-annealing rig: The samples were mounted on

a homemade test fixture (Figure S1), consisting of

aluminum foil sandwiched between a PCB circuit

board and a 1 mm-thick glass slide. Each sample

was placed on top of the glass slide and secured

with two SEM clips (PELCO SEMClip) that were

attached to the circuit board with screws. Constant

current was injected into the sample through the

clips. We used a switching DC Power supply 1685B

(BK Precision) to set constant current and monitor

the voltage across the sample. Temperature of the

thin film was directly measured by an infrared (IR)

thermometer (thermoMETER LS by Micro-Epsilon)

stably-positioned over the sample (Figure S-2 in the

ESM). The IR spot was positioned at the middle of a

trace for unconstricted samples and at the

constriction region for the constricted samples.

Note that the minimum spot size of the IR

thermometer is 1 mm, thus it was not possible to


4 Nano Res.

measure temperature for the constriction sizes

smaller than the spot size.

IR thermometers use emissivity to convert detected

infrared radiation to temperature. We found that an

emissivity value of 0.17 gives reasonably accurate

temperature measurements of np-Au surface,

covering the entire range of 20 ˚C to 300 ˚C used in

this study (Figure S-3b in the ESM). Briefly, the

calibration procedure involved the two

thermocouples attached onto the surface of hot

plate and the IR thermometer positioned above the

sample (Figure S-3a in the ESM). Temperatures

were recorded every 1 second. Emissivity for np-Au

was calibrated by measuring the temperature of the

np-Au thin film sample on the hot plate via IR

thermometer and comparing it to the temperature

readings of both thermocouples. This procedure

was repeated for temperature settings on the hot

plate from 150 ˚C to 300 ˚C in 25 ˚C increments.

Each temperature was held for 2 minutes and a

fresh sample was used for each distinct temperature

setpoint. In an additional experiment, a fresh

sample was placed on hot plate at setpoint of 75 ˚C

for 2 minutes; temperature setpoint was

incremented in 25 ˚C intervals and held for 2

minutes until 300 ˚C setpoint. After this cycle, the

annealed sample was allowed to cool down to room

temperature and the calibration procedure was

repeated with the same annealed sample to capture

possible effects of coarsening on IR-based

measurement in comparison to the thermocouple

references. For the an emissivity setting of 0.17,

between 150 ˚C and 300 C, a measurement error of

IR thermometer was less than 7% as compared to

the temperature obtained by thermocouples. This

error was less than 15% for temperatures between

75 C and 150 C (Figure S-3B in the ESM).

Scanning electron microscopy and morphology analysis:

Top and cross-sectional views of the samples were

obtained with a scanning electron microscope (FEI

Nova NanoSEM430) at 100 kX magnification. The

film thickness was determined by examining the

cross-sectional images. Highly-coarsened

morphologies resulted in few ligaments in a single

image at this magnification; therefore for such

samples additional images at a lower magnification

were obtained. The image analysis results from the

50 kX and 100 kX were pooled together to provide a

representative ligament size. SEM images were

analyzed using ImageJ (National Institutes of

Health shareware). For consistency, automatic

thresholding option, which is a variation of IsoData

algorithm [32], in ImageJ was used to convert

grayscale images to monochrome, where pores are

displayed as black and ligaments as white regions.

We used MatLab to calculate ligament thickness

from the monochrome images. Black and white

images were scanned line by line, where ligaments

were detected by transition from a black pixel to a

white pixel and the number of pixels in each

ligament (white pixels) was determined. The total

number of pixels for each ligament was converted

to a metric value using the SEM scale bars and the

median ligament thicknesses were extracted for

each image. The effective surface area of the np-Au

samples were determined by electrochemical oxide

stripping in dilute sulfuric acid as previously

described [20].

3. Results and Discussion

3.1 Effect of electro-annealing duration on

morphology evolution

We initially studied how pore morphology varies

with electro-annealing duration using the

unconstricted rectangular np-Au traces (Figure 1A).

We evaluated different current settings and

Figure 1. Unconstricted np-Au samples. (a) Photographs of samples that have undergone electro-annealing for 60, 67, 75, 82 and 100

s; and (b) corresponding SEM images taken at the center of each trace. (c) Median ligament thickness vs x-position for each sample.

The fitting lines are for visual guides only. Inset shows the ligament thickness versus y-position at x = 11 mm for the 82 s-sample.

The extent and span of coarsening increases with electro-annealing duration. (d) Definition of the coordinate nomenclature and the

heat map illustrating the temperature distribution over the trace.

determined that 1.5 A offered desirable control over

ligament coarsening, as evidenced by color change

visible to naked eye. In contrast, lower current

settings resulted in negligible coarsening, while

higher currents resulted in erratic coarsening.

Figure 1a and 1b show photographs and

corresponding SEM images of the samples treated

at different electro-annealing durations. Figure 2

illustrates cross-sectional SEM images of selected

samples that capture the range of morphology

evolution shown in Figure 1b. No significant

change in pore morphology was observed for

treatment durations less than 60 seconds. The color

change in the samples treated for longer than 60

seconds indicates higher temperatures in those

areas, which is in agreement with coarser pore

morphology observed in the SEM images. These

high-temperature hot spots generally appear in the

middle of the trace, implying that temperature

distribution is not uniform over the sample surface

(to be discussed below).

To study the hot spot evolution further, we

determined median ligament thickness for each

sample along the trace length (x-direction) in 2 mm


6 Nano Res.

steps (Figure 1c). The untreated sample had a

median ligament thickness of 58 nm. For short

treatment durations, the ligament size did not

change significantly (i.e., median ligament size for

the 60-second treatment is 73 nm). However, the

ligament size increased dramatically to

approximately 200 nm for the 100 second treatment

duration. For this duration, the pore morphology

deviated from the characteristic interconnected pore

morphology at several locations on the trace and

exhibited a more columnar structure uncovering the

substrate surface below the porous film (see SEM

for 100 s in Figure 1b). The pore morphology

through the film thickness remains uniform for

moderate electro-annealing (Figure 2b). In the case

of heavy-coarsening (i.e., 100 s duration in Figure

1b), the ligaments coalesce into thick columnar

structures (Figure 2c) and the number of pores

decreases significantly. These cross-sectional

images are also consistent with those of typical

morphologies obtained by thermal treatment. In

addition to the initiation of ligament thickening at

the middle of the trace, the coarsening process

progresses outwards with increasing treatment

duration. This observation is evident for samples

treated at 75, 82 and 100 seconds with the latter

exhibiting a coarsened morphology in nearly the

entire length of the trace (Figure 1a & 1c). We also

imaged the 82 seconds sample along the width

(y-direction) at the middle of the trace, (inset Figure

1c). In accord with the morphology profile along

the trace length, the ligament coarsening is more

pronounced in the middle of the trace and drops off

slightly near the edges.

The maximum temperature measured at the middle

of the trace for the 1.5 A current application for all

treatment durations was 140 ± 5 ˚C. It should be

noted that this average value excluded the

temperature (172 ˚C) for the 100 s case because of its

atypical morphology (Figure 1b and Figure 2c) and

likely effect on electro-annealing behavior.

Figure 2. Cross-sectional SEM images of electro-annealed

np-Au samples in Figure 1. The images taken at the hotspot

correspond to treatment durations of (a) control, (b) 75 s, and

(c) 100 s.

Variation of ligament size across the length of the

sample suggests that the highest temperature

occurs at the middle of the trace. To investigate the

temperature distribution over the sample, we

measured temperature as a function of position on a

5 mm wide trace by rastering the IR thermometer.

We specifically used a wider trace to allow for (i)

non-overlapping temperature measurements (note

the 1 mm IR spot size); and (ii) a lower maximum

temperature at 1.5 A due to a larger trace

cross-section and thus less electro-annealing to

maintain a stable surface for accurate

surface-temperature interrogation. The heat map

indeed indicated that the highest temperature

coincides with the middle of the trace (Figure 1d),

consistent with the ligament coarsening results

(Figure 1c). The observed temperature gradient

along the trace offers a unique opportunity to create

gradually-varying pore morphologies over several

millimeters.

The effective surface area of a coating plays an

important role for applications ranging from

catalysts [16, 17] to biomedical coatings [9-12]. To

that end, we tracked changes in surface area at

various stages of pore coarsening using

electrochemical oxide stripping protocol described

previously [20]. While the varying pore

morphology of electro-annealed films makes it

difficult to obtain an exact surface area that

corresponds to a specific pore morphology, we


7 Nano Res.

quantified the surface area of thermally-annealed

np-Au films with a uniform pore morphology.

These samples were annealed on the hot plate at

temperatures that range between 225 °C to 275 °C

and capture the range of morphologies obtained via

electro-annealing (Figure S-4). The increase in

effective surface area with respect to that of a planar

gold surface with identical dimensions is referred to

as enhancement factor, which is expected to decrease

with increasing feature size. While the un-annealed

np-Au film has a surface area enhancement of 12.8,

the 225 ˚C sample (morphology corresponding to 67

s treatment) had an enhancement factor of 10.7. For

the heavily-coarsened morphology (similar to that

observed for the 100 s case in Figure 1), the

enhancement factor was reduced to 5.1. The drastic

change in enhancement factor for samples annealed

at temperatures only 25 ˚C apart highlights the

importance of precise control over pore

morphology enabled by electro-annealing.

3.2. Effect of geometric constraints on morphology

evolution

The spatial location of the hot spot can be controlled

by introducing a lithographically-defined geometric

constraint in the trace, as shown in the Figure 2a.

The dimensions of the constrictions are 2 x 2 mm

and overall dimension of the trace is 4 x 24 mm.

Basic circuit theory suggests that the geometric

constriction causes increased current density in the

constriction region and leads to localized

coarsening in that area as compared to the rest of

the trace. By introducing the constriction at

different positions along the trace, we demonstrated

that it is possible to control the location of the hot

spot and even steer it to multiple and arbitrary

locations (Figure 3a).

The SEM images from the middle of constrictions

and in the adjacent areas indicate that the

constrictions not only allow for controlling the

location of the hot spot but also the span of the

morphology gradient centered around the hot spot.

This finding is illustrated via superimposing the

ligament size variations along two different trace

geometries (i.e., unconstricted and with one

constriction) and aligning the peak locations, as

shown in Figure 3c. For the unconstricted case, the

coarsened morphology spans approximately 10 mm

and shows gradual variation in ligament size with

maximum size approaching 210 nm in the middle

of the trace. In contrast, for the constricted case

coarsening only spans 3 mm with maximum

ligament size of approximately 215 nm, a sharp

slope in ligament versus distance plot for this

sample demonstrates a well-controlled transition

between the annealed and unannealed regions of

the sample. The ability to precisely control the

location and span of pore morphology opens the

door to producing high-throughput platforms for

studying property-structure relationships.

We hypothesized that having a design with various

constriction dimensions on a single trace would

create the possibility of engineering a chip with

multiple pore morphologies. One such design

(labeled as “iv”) is shown in Figure 3a. In this

configuration, we designed the trace with three

progressively decreasing constriction widths (i.e., 2

mm, 1.875 mm, and 1.75 mm). The results indicate

that majority of the coarsening occurs in the middle

despite the thinnest constriction being on the right.

This observation is due to uneven heat dissipation

and consequent temperature distribution across the

sample as previously discussed (Figure 1d). These

results suggest that constriction position and

dimensions should be carefully engineered to

compensate for higher temperatures at the middle

of the trace.

3.3. Branched structures for producing multiple

pore morphologies

An alternative design approach with more degrees

of freedom is to have multiple traces connected in

Figure 3. Electro-annealing of samples with geometric constraints. (a) Photos of unconstricted and constricted samples. All

constrictions are 2 mm in length (x-direction) and the constriction widths (y-direction) are indicated on the images. (b) SEM images

at the constriction and periphery are denoted with the white spots on the photographs in the insets. Trace constrictions allow for

programming the location of pore coarsening. (c) Ligament thickness versus position for constricted (2 mm x 2 mm) and

unconstricted (annealed for 100 s) samples. The curves were aligned at the peak ligament size for ease of visual comparison

parallel, with constrictions of various dimensions in

the middle of each trace (Figure 4a). The major

advantage of this approach is that the current

density in each branch can be individually

controlled not only by the width of the constriction

but also by the width of each main trace. This

approach can be scaled up to multiple branches

informed by electrical circuit (current divider)

principles, thus providing a higher degree of

control over the resulting pore morphology in each

branch. Here, we investigated a basic circuit

topology, where each main trace dimension is

identical while the constriction dimensions vary.

The rationale for this design is that narrower

constrictions experience higher Joule heating and

enhanced electrically-assisted mechanisms (leading

to coarser morphologies) as the current density

increases with decreasing constriction width. The

constriction lengths were all 1 mm, while the width

varied between 0.5 mm to 0.9 mm across the five

different traces (Figure 4a).

As a basic circuit analysis would reveal, the current

in each branch is largely dictated by the dimensions

of the main line, as the constriction is significantly

less resistive compared to the main line due to its

reduced length. For an applied current of 2 A, we

expect the current through each trace to be on the

order of 400 mA. Taking constrictions into account,

the currents between two successive branches are

expected to differ by less than 9 mA. Even though

the resistance of the constriction regions increases

with temperature, the overall resistance is still

largely determined by the main line, as such the

current within each branch should not change by

more than 7% during electro-annealing. As

supported by the SEM images at each constriction

(Figure 4b), except for Line 5, a coarser morphology

is evident as constrictions become narrower. While

Line 5 with the narrowest constriction was expected

to have the coarsest morphology due to the highest

current density (Figure 4c), the ligament size was

between those for Lines 2 and 3 (Figure 4c). This

observation is again in agreement with previously

discussed uneven heat dissipation and temperature

differentials across the sample (Figure 1d). Put

another way, the inner lines (2, 3, and 4) run

warmer as they receive heat dissipated from the

traces around them, while the outer lines (1 and 5)

Figure 4. Branched structures for differential electro-annealing. (a) The branched trace design and corresponding SEM images of

each constriction in (b). (c) Median ligament thickness and a current-geometry index (i.e., current through each main line divided by

individual constriction widths) plotted for each trace. The increasing current-geometry index largely correlates with the extent of

ligament coarsening, excluding peripheral lines that are subject to more rapid heat loss.

lose heat faster leading to lower temperatures.

In order to further decouple the effect of varying

constriction geometry from the non-uniform

temperature distribution, we employed a

parallel-trace configuration similar to that used in

Figure 4a, except each trace had three identical

constrictions. When this trace network was

electro-annealed, the coarsening was localized at

the constrictions and the influence of temperature

distribution (observed in Figure 1d) was

superimposed onto the localized coarsening (Figure

S-5b in the ESM). More specifically, the constraints

at the center were the most coarsened and the

coarsening at the constrictions decreased radially

outward from the center. This emphasizes that

temperature gradient across the sample alone can

be utilized for producing varying pore

morphologies localized at the constrictions.

However, in order to realize a large number of

controllable morphologies on a single chip with

high spatial and morphological precision, it is

necessary to optimize geometric constrictions by

incorporating the effect of uneven heat dissipation

into the trace circuit design.

3.4. Comparison of electro-annealing and thermal

treatment

The effective temperatures measured during

electro-annealing were significantly lower than

annealing temperatures reported in literature for

thermal treatment [23, 29, 31, 33, 34]. Comparison

of ligament sizes and temperatures that we

obtained during electro-annealing and hot plate

calibration also indicated that much higher

temperatures were required during thermal

annealing to obtain morphologies comparable to

those observed with electro-annealing. To confirm

this observation, we performed an electro-annealing

experiment using an unconstricted 3.5 mm-wide


10 Nano Res.

np-Au trace and measured a maximum

temperature of approximately 155 ˚C in the middle

of the strip after the color change in the sample

(indicative of substantial coarsening). The duration

of this treatment was three minutes and the

resulting median ligament thickness at the hot spot

was 156 nm (Figure 5a left). However, when a

np-Au sample (identical to that used for the

electro-annealing test) was treated at 155 ˚C

(temperature monitored with the same IR

thermometer) on the hot plate, there was no color

change after 3 minutes of annealing and the

measured ligament thickness was 61 nm (Figure 5a

right) – similar to that of unannealed samples (58

nm).

Based on hot plate calibration data, it is expected to

observe significant coarsening only after 250 ˚C.

Indeed, no major coarsening or color change was

observed for 3 minute-treatment on the hot plate for

temperatures up to 220 ˚C (Figure 5b left). At 250 ˚C

on the hot plate, it took 1.5 minutes for np-Au

sample to change color. This suggests that in

order to obtain comparable ligament sizes between

electro-annealing (Figure 5a left) and the hot plate

(Figure 5b right), the np-Au temperature on the hot

plate should be between 220 ˚C and 250 ˚C, which is

nearly 100 ˚C more than effective temperatures

recorded for electro-annealing. This hints that

temperature alone may not be the only mechanism

responsible for ligament coarsening during

electro-annealing. The coarsening of np-Au is

primarily driven by surface-diffusion-based

processes [29, 35]. The surface diffusion coefficient

that dictates the rate of the coarsening process

follows an Arrhenius relationship [29]:

𝐷 = 𝐷0exp(−𝐸𝐴

𝑘𝑇) (1)

where D0 is the pre-exponential factor, EA is the

activation energy, T is the temperature, and k is the

Boltzmann constant.

Figure 5. Morphology and temperature comparison for

electro-annealing (EA) and hot plate (HP). (a) SEM images of

two identical np-Au samples treated at 155 ˚C for 3 minutes via

electro-annealing (left) and on hot plate (right). Note that for

the same treatment temperature and duration, electro-annealing

resulted in significant coarsening while no coarsening was

observed for the hot plate case. (b) In order to obtain a similar

degree of coarsening with electro-annealing case shown in part

(a), a hot plate treatment at 250 ˚C for 1.5 minutes was

necessary (right). Note that a 3-minute treatment on hot plate at

an intermediate temperature of 220 ˚C did not match the

coarsening observed for electro-annealing.

As the equation suggests, the diffusion coefficient,

D, increases with temperature, and leads to the

experimentally-observed coarsening [31, 33, 34]. For

traditional thermal treatment approach, the

activation energy necessary to initiate significant

surface diffusion is largely provided by thermal

energy (kT term) alone. The reduced temperature

for the diffusion process, observed during

electro-annealing, suggests that additional energy

sources complement the thermal energy in

initiating the surface diffusion. High current

densities have been known to result in

electromigration of atoms within wires, which

ultimately leads to electromigration-led failure. The


11 Nano Res.

critical current densities for this failure mechanism

in gold nanowires with similar characteristic

length-scale to that for the np-Au ligaments, has

been reported to be on the order of 1012 A/m2. This

critical current density is three orders of magnitude

higher than the estimated current density per

np-Au ligament (on the order of 109 A/m2) for the

unconstricted traces investigated here (Figure 1).

Also, the SEM images of the electro-annealed

samples did not reveal any evidence of

electromigration, such as material accumulation

downstream of the electron flow direction or

nanogap formation in ligaments [36, 37]. It is

widely-reported that current flow through small

structures (e.g., thin films, nanowires) where mean

free path of electrons become comparable to

characteristic length-scales of the conductor, the

electronic transport is dominated by

electron-surface scattering [38, 39]. In the context of

np-Au thin films, which can be visualized as a 3D

network for nanowires (ligaments), the surface

corresponds to the ligament-nanopore interface [38].

Indeed, these scattering events and the resulting

energy loss manifest themselves as higher electrical

resistance in thin films compared to their bulk

counter-part and may be a source of additional

energy assisting the coarsening process [40].

For a qualitative discussion of the electro-annealing

case, it is helpful to conceptualize the effect of

additional energies by modifying the denominator

term of the exponential in the Arrhenius equation to

kT + Eother, where Eother signifies other putative

energies. Some kinetic energy during

electro-annealing is dissipated as heat (kT term)

which is the temperature rise observed during

electro-annealing. The electron collisions during the

scattering events would apply a force on the

adatoms and the electric field in the ligaments (due

to the applied voltage) would exert an additional

force on the adatoms. These two forces would

constitute Eother and translate into increasing the

average energy per adatom. Consequently, the

surface diffusion of adatoms (thus coarsening) may

occur with reduced contribution from the kT term.

It is also plausible that the surface diffusion occurs

before electromigration, implying that the observed

phenomenon may be happening at a transition

regime between thermal diffusion and

electromigration. The net result is that coarsening

occurs at lower temperatures for electro-annealing

than for pure thermal treatment. Finally, it should

be noted that the complex interplay between these

processes that constitute Eother are beyond the scope

of this work and require further studies, including

molecular simulations

4. Conclusion

By taking advantage of the electrical conductivity

and morphological plasticity of np-Au, we

demonstrated that electro-annealing is a novel and

versatile technique to controllably tune pore

morphology of np-Au thin films at low

temperatures. More specifically, we introduced

geometric constrictions into

lithographically-patterned np-Au traces and

obtained distinct pore morphologies with

controllable location and span. This technique is

broadly applicable to tuning the nanostructure of

other electrically-conductive nanoporous metal

systems [6] and amenable to further

functionalization via conformal atomic layer

deposition of ceramics [41]. We expect this novel

approach to be a path to programmable

morphology generation using electrical circuit

principles and associated CAD tools to design

complex trace networks and assist high-throughput

screening of structure-property relationships for

fundamental and applied studies. We envision that

further work in developing a design toolkit that

generates optimized trace geometries and

electro-annealing parameters will allow for

precisely controlling the location and extent of


12 Nano Res.

coarsening. In contrast to traditional thermal

coarsening techniques (e.g., furnace, rapid thermal

annealing), electro-annealing allows to selectively

modify pore morphology simultaneously over

different regions on a substrate. When compared to

other nascent annealing techniques (e.g., laser

sintering [29], the electro-annealing technique does

not require sophisticated optical instrumentation,

thereby making the method generally available to

users with access to simple photolithography and

deposition capabilities. While laser micro sintering

offers high spatial resolution due to the small laser

spot size (few microns), it remains unpractical for

creating larger structures (100s of microns to

millimeters). In addition, we also observed that

temperatures required to obtain significant

ligament coarsening during electro-annealing are

much lower than temperatures used during

traditional thermal annealing, suggesting that other

electrically-assisted mechanisms contribute or

enhance ligament coarsening during

electro-annealing.

Acknowledgements

We gratefully acknowledge the support from UC

Lab Fees Research Program Award (12-LR-237197),

Research Investments in the Sciences & Engineering

(RISE) Award, and UC Davis College of

Engineering start-up funds.

Electronic Supplementary Material: Supplementary

material further details of the fabrication, IR

thermometer calibration, and setup scheme is

available in the online version of this article at

http://dx.doi.org/10.1007/s12274-***-****-*

(automatically inserted by the publisher).

References [1] Gleiter, H. Nanostructured materials: Basic

concepts and microstructure. Acta Materialia 2000,

48, 1-29.

[2] Suryanarayana, C. Recent developments in

nanostructured materials. Advanced Engineering

Materials 2005, 7, 983-992.

[3] Zhang, X.; Xie, H.; Fujii, M.; Ago, H.;

Takahashi, K.; Ikuta, T.; Abe, H.; Shimizu, T.

Thermal and electrical conductivity of a suspended

platinum nanofilm. Applied Physics Letters 2005, 86,

171912.

[4] Balandin, A. A. Thermal properties of

graphene and nanostructured carbon materials. Nat

Mater 2011, 10, 569-581.

[5] Konstantatos, G.; Sargent, E. H.

Nanostructured materials for photon detection. Nat

Nano 2010, 5, 391-400.

[6] Seker, E.; Reed, M.; Begley, M. Nanoporous

gold: Fabrication, characterization, and applications.

Materials 2009, 2, 2188-2215.

[7] Xiao, X.; Wang, M. e.; Li, H.; Si, P. One-step

fabrication of bio-functionalized nanoporous

gold/poly(3,4-ethylenedioxythiophene) hybrid

electrodes for amperometric glucose sensing.

Talanta 2013, 116, 1054-1059.

[8] Hu, K.; Lan, D.; Li, X.; Zhang, S.

Electrochemical DNA biosensor based on

nanoporous gold electrode and multifunctional

encoded DNA−au bio bar codes. Analytical

Chemistry 2008, 80, 9124-9130.

[9] Seker, E.; Berdichevsky, Y.; Begley, M.; Reed,

M.; Staley, K.; Yarmush, M. The fabrication of

low-impedance nanoporous gold multiple-electrode

arrays for neural electrophysiology studies.

Nanotechnology 2010, 21, 125504.

[10] Seker, E.; Berdichevsky, Y.; Staley, K. J.;

Yarmush, M. L. Microfabrication-compatible

nanoporous gold foams as biomaterials for drug

delivery. Advanced Healthcare Materials 2012, 1, 172–

176.

[11] Tan, Y. H.; Schallom, J. R.; Ganesh, N. V.;

Fujikawa, K.; Demchenko, A. V.; Stine, K. J.

Characterization of protein immobilization on

nanoporous gold using atomic force microscopy

http://dx.doi.org/10.1007/s12274-***-****-*


13 Nano Res.

and scanning electron microscopy. Nanoscale 2011, 3,

3395-3407.

[12] Patel, J.; Radhakrishnan, L.; Zhao, B.;

Uppalapati, B.; Daniels, R. C.; Ward, K. R.;

Collinson, M. M. Electrochemical properties of

nanostructured porous gold electrodes in

biofouling solutions. Analytical chemistry 2013, 85,

11610-11618.

[13] Lang, X.; Hirata, A.; Fujita, T.; Chen, M.

Three-dimensional hierarchical nanoporosity for

ultrahigh power and excellent cyclability of

electrochemical pseudocapacitors. Advanced Energy

Materials 2014, 4, 1301809.

[14] Kucheyev, S.; Hayes, J.; Biener, J.; Huser, T.;

Talley, C.; Hamza, A. Surface-enhanced raman

scattering on nanoporous au. Applied Physics Letters

2006, 89, 053102-053104.

[15] Santos, G. M.; Zhao, F.; Zeng, J.; Shih, W.-C.

Characterization of nanoporous gold disks for

photothermal light harvesting and light-gated

molecular release. Nanoscale 2014, 6, 5718-5724.

[16] Wittstock, A.; Zielasek, V.; Biener, J.; Friend, C.

M.; Bäumer, M. Nanoporous gold catalysts for

selective gas-phase oxidative coupling of methanol

at low temperature. Science 2010, 327, 319-322.

[17] Xu, C.; Su, J.; Xu, X.; Liu, P.; Zhao, H.; Tian, F.;

Ding, Y. Low temperature co oxidation over

unsupported nanoporous gold. Journal of the

American Chemical Society 2006, 129, 42-43.

[18] Lee, D.; Wei, X.; Chen, X.; Zhao, M.; Jun, S.;

Hone, J.; Herbert, E.; Oliver, W.; Kysar, J.

Microfabrication and mechanical properties of

nanoporous gold at the nanoscale. Scripta materialia

2007, 56, 437-440.

[19] Jin, H. J.; Weissmüller, J. A material with

electrically tunable strength and flow stress. Science

2011, 332, 1179-1182.

[20] Kurtulus, O.; Daggumati, P.; Seker, E.

Molecular release from patterned nanoporous gold

thin films. Nanoscale 2014, 6, 7062 - 7071.

[21] Senior, N.; Newman, R. Synthesis of tough

nanoporous metals by controlled electrolytic

dealloying. Nanotechnology 2006, 17, 2311-2316.

[22] Erlebacher, J. An atomistic description of

dealloying. Journal of the Electrochemical Society 2004,

151, C614-C626.

[23] Hakamada, M.; Mabuchi, M. Thermal

coarsening of nanoporous gold: Melting or

recrystallization. Journal of Materials Research 2009,

24, 301-304.

[24] Fujita, T.; Qian, L.-H.; Inoke, K.; Erlebacher, J.;

Chen, M.-W. Three-dimensional morphology of

nanoporous gold. Applied Physics Letters 2008, 92,

251902.

[25] Qian, L. H.; Chen, M. W. Ultrafine

nanoporous gold by low-temperature dealloying

and kinetics of nanopore formation. Applied Physics

Letters 2007, 91, 083105.

[26] Detsi, E.; van de Schootbrugge, M.; Punzhin,

S.; Onck, P. R.; De Hosson, J. T. M. On tuning the

morphology of nanoporous gold. Scripta Materialia

2011, 64, 319-322.

[27] Dong, H.; Cao, X. Nanoporous gold thin film:

Fabrication, structure evolution, and electrocatalytic

activity. The Journal of Physical Chemistry C 2008, 113,

603-609.

[28] Hakamada, M.; Mabuchi, M. Microstructural

evolution in nanoporous gold by thermal and acid

treatments. Materials Letters 2008, 62, 483-486.

[29] Schade, L.; Franzka, S.; Mathieu, M.; Biener,

M. M.; Biener, J.; Hartmann, N. Photothermal laser

microsintering of nanoporous gold. Langmuir 2014,

30, 7190-7197.

[30] Klein, R. Laser welding of plastics. Wiley: 2011.

[31] Daggumati, P.; Kurtulus, O.; Chapman, C. A.

R.; Dimlioglu, D.; Seker, E. Microfabrication of

nanoporous gold patterns for cell-material

interaction studies. 2013, e50678.

[32] Sezgin, M.; Sankur, B. Survey over image

thresholding techniques and quantitative

performance evaluation. Journal of Electronic Imaging

2004, 13, 146-168.

[33] Hodge, A. M.; Biener, J.; Hayes, J. R.; Bythrow,

P. M.; Volkert, C. A.; Hamza, A. V. Scaling equation


14 Nano Res.

for yield strength of nanoporous open-cell foams.

Acta Materialia 2007, 55, 1343-1349.

[34] Li, R.; Sieradzki, K. Ductile-brittle transition

in random porous au. Physical Review Letters 1992,

68, 1168-1171.

[35] Erlebacher, J. Mechanism of coarsening and

bubble formation in high-genus nanoporous metals.

Physical Review Letters 2011, 106, 225504.

[36] Trouwborst, M. L.; van der Molen, S. J.; van

Wees, B. J. The role of joule heating in the formation

of nanogaps by electromigration. Journal of Applied

Physics 2006, 99, 114316.

[37] Hadeed, F. O.; Durkan, C. Controlled

fabrication of 1–2nm nanogaps by electromigration

in gold and gold-palladium nanowires. Applied

Physics Letters 2007, 91, 123120.

[38] Fujita, T.; Okada, H.; Koyama, K.; Watanabe,

K.; Maekawa, S.; Chen, M. W. Unusually small

electrical resistance of three-dimensional

nanoporous gold in external magnetic fields.

Physical Review Letters 2008, 101, 166601.

[39] Liu, Z.; Searson, P. C. Single nanoporous gold

nanowire sensors. The Journal of Physical Chemistry B

2006, 110, 4318-4322.

[40] Munoz, R. C. Resistivity induced by a rough

surface of thin gold films deposited on mica. J. Mol.

Catal. A-Chem. 2005, 228, 163-175.

[41] Biener, M. M.; Biener, J.; Wichmann, A.;

Wittstock, A.; Baumann, T. F.; Baumer, M.; Hamza,

A. V. Ald functionalized nanoporous gold: Thermal

stability, mechanical properties, and catalytic

activity. Nano Letters 2011, 11, 3085-3090.


Nano Res.

Electronic Supplementary Material


gold thin

Tatiana S. Dorofeeva1, and Erkin Seker1,2()

Supporting information to DOI 10.1007/s12274-****-****-* (automatically inserted by the publisher)

Sample Preparation

The glass cover slips used as substrates, on which metal traces are to be patterned, were cleaned by

immersion in a piranha solution for 7 minutes, rinsed in deionized (DI) water, and dried under nitrogen flow.

In order to micropattern the metal traces, the cover slips were coated with ~1 µm thick positive photoresist

(Shipley 1813) followed by a pre-bake for 90 seconds at 115 ˚C on a hot plate. Transparency mask was used

for photolithography. Alignment and exposure was performed on Karl Suss mask aligner with exposure

duration of 4.5 seconds at 20 mJ. Exposure was followed by a post-bake for 90 seconds at 115 ˚C. The exposed

photoresist was developed in MF-322 (Shipley) for 25 seconds with slight manual agitation. The samples

were subsequently loaded into a sputtering machine (Kurt J. Lesker) for thin film deposition. Initially, 120 nm

of chrome (to promote adhesion between glass and gold) was sputtered at 300 W, followed by 80 nm of seed

gold at 400 W, finally 500 nm of silver and gold were co-sputtered at 100 W and 200 W respectively. All

depositions were performed under argon atmosphere at 10 mTorr. The lift-off was performed in NMP

(N-methylpyrrolidone) by cycles of 10 seconds of sonication followed by a 10 minute pause until all the

photoresist was removed. Elemental compositions of the samples were ascertained with energy dispersive

X-ray spectroscopy (Oxford INCA Energy-EDS), the resulting gold and silver content of the films was 36%

Au and 64% Ag by atomic %. Deposited samples were dealloyed in 70% nitric acid for 15 minutes at 55 ˚C to

produce the np-Au films. The dealloying process typically leads to 3-5% residual silver in the final np-Au

film. To completely remove traces of nitric acid after the dealloying, samples were rinsed several times and

stored in DI water. After 2 days, samples were removed from water, dried with nitrogen gun and stored in

ambient conditions until the conducting the electro-annealing experiments.

Measurement Setup

Samples were mounted on the homemade electro-annealing test fixture (Figure S-1), consisting of an

aluminum foil sandwiched between a PCB circuit board and a 1 mm-thick glass slide. The sample was placed

on top of glass slide and secured with two SEM clips (PELCO SEMClip) which were attached to the circuit

board with screws. Constant current was injected into the sample through the clips. We used switching DC

————————————



Nano Res.

Power supply 1685B (BK Precision) to set constant current and monitor the voltage across the sample.

Temperature was measured by an infrared thermometer (thermoMETER LS by Micro-Epsilon)

stably-positioned over the sample (Figure S-2). For the unconstricted samples the temperature was measured

in the middle of the trace, for the constricted samples temperature was measured at the constrictions.

Figure S-1. Electro-annealing test fixture


Nano Res.

Figure S-2. The setup for electro-annealing is shown as well as sample voltage, current and temperature readings taken during

experiments.


Nano Res.

Temperature Measurements

Infrared thermometers use emissivity to convert detected infrared radiation to temperature, thus properly

choosing the emissivity of np-Au is essential for obtaining proper temperature readings of the np-Au thin

films. We utilized two thermocouples (type K), a separate digital thermometer (HH802U by Omega) and a

hot plate to calibrate the infrared (IR) thermometer emissivity setting for np-Au. Two thermocouples were

attached to the surface of the hot plate using a Kapton tape. One thermocouple was connected to the digital

thermometer and second was connected to an auxiliary input port of the IR thermometer. The IR

thermometer was positioned above the sample and focused using its laser alignment marks. Both

thermometers were connected to a computer through USB cables and the measured temperatures were

recorded every 1 second.

The emissivity for np-Au was calibrated by measuring the temperature of the np-Au thin film sample on the

hot plate and comparing it to the temperature readings of both thermocouples. IR thermometer calibration in

Figure S-3b shows temperature of the hot plate measured using two thermocouples (TCΩ and TCIR) and

temperature of np-Au (TIRnp-Au) measured by IR thermometer with emissivity set to 0.17. This procedure was

repeated for temperature settings on the hot plate from 150 ˚C to 300 ˚C in 25 ˚C increments, each

temperature was held for 2 minutes. Fresh np-Au sample was placed on the hot plate before moving on to

new setpoint. The calibration indicated that an emissivity (ε) setting of 0.17 gives reasonably accurate

temperature measurements of np-Au surface over an entire temperature range 150 ˚C to 300 ˚C. Figure S -3b

shows that the temperatures measured by the two thermocouples and temperature of np-Au are all within 10

˚C of each other in 150 to 300 C.

As an additional experiment to determine if changing morphology during coarsening impacts the accuracy

of IR temperature measurement, we performed a similar experiment with only one sample in an expanded

lower temperature range from 75 to 300 ˚C. More specifically, a fresh sample was placed on hot plate at

setpoint of 75 ˚C for 2 minutes; temperature setpoint was incremented in 25 ˚C intervals and held for 2

minutes until 300 ˚C setpoint. After this cycle, the annealed sample was allowed to cool down to room

temperature and the calibration procedure was repeated with the same annealed sample to capture possible

effects of coarsening on IR-based measurement in comparison to the thermocouple references. No significant

deviations (< 15 ˚C for 75-150 ˚C and < 9 ˚C for 150-300 ˚C) between IR thermometers and thermocouples

were observed (Figure S-3b) for the samples covering the entire range of morphologies produced in this

study. This suggests that the emissivity of np-Au did not change appreciably at least for the studied

morphologies within investigated temperature range.


Nano Res.

Figure S-3. (a) Temperature calibration setup. (b) The IR thermometer (emissivity set to 0.17) calibration curves with respect to

thermocouples for np-Au samples with different degrees of coarsening. The TIR and TC abbreviations in the legend indicate the

temperature readings that were obtained by the infrared thermometer and the thermocouples respectively. TIRnp-Au n/anneal : An unannealed

sample was placed on the hot plate and temperature was increased in 25 °C increments from 75 °C to 300 °C while acquiring the sample

temperature following two minutes of stabilization after each setpoint. TIRnp-Au annealed: The same temperature setpoint and acquisition

protocol from the previous experiment was used. However, this time an annealed sample (produced in the TIRnp-Au n/anneal experiment)

was used to evaluate the effect of an already coarsened morphology on IR thermometer readings. TIRnp-Au new: A fresh unannealed

sample was used for each temperature point between 150 °C and 300 °C in 25 °C increments to evaluate the effect of temperature

exposure on coarsening at different temperatures. TCIR: Temperature of hot plate directly adjacent to the sample obtained by the

thermocouple attached to IR thermometer. TCΩ: Temperature of hot plate directly adjacent to the sample obtained by the thermocouple

attached to an additional thermometer.

Surface Enhancement Factor

Electrochemical oxide stripping technique was employed to determine the effective surface areas of np-Au

samples. The electrochemical surface characterization (enabled by the electrical conductivity of np-Au) is

based on the principle that AuO reduction only happens at the gold surface in contact with the electrolyte

and the surface reaction-limited process allows for addressing the entire surface of the np-Au thin film.

While the varying pore morphology of electro-annealed films makes it difficult to obtain an exact surface

area that corresponds to specific pore morphology, we quantified the surface area of thermally-annealed

np-Au films with uniform pore morphology. Fresh np-Au samples were thermally-annealed on the hot plate

at 225 °C, 250 °C, and 275 °C for 2 minutes to capture the range of morphologies obtained via

electro-annealing. Each np-Au sample with a unique morphology was loaded into a custom-built

three-electrode Teflon electrochemical cell with platinum counter electrode and Au/AgCl reference electrode.

The np-Au sample was immersed in 0.05 M sulfuric acid and scanned at a rate of 50 mV/s to obtain the

corresponding cyclic voltammogram. The total electrical charge under gold oxide reduction peak between

potentials of 720 mV and 970 mV was converted to surface area using 450 μC/cm2 as the specific charge of a

gold surface [1, 2]. The ratio of the effective surface area of the measured sample to that of a planar gold

electrode with the same foot print was referred to as “enhancement factor” In Figure S -4 below surface

enhancement factor for four different temperature annealing ranges are plotted along with corresponding

SEM images illustrating the surface morphology representative for thermal annealing at this temperature,

the size of each SEM image is approximately 1x1 µm, we estimate that corresponding JH durations are 67 s,

82 s, and 100 s respectively.


Nano Res.

Figure S-4. Area enhancement factor for thermally annealed npAu samples at temperatures of 225 °C, 250 °C, and 275 °C. As guide

only, the corresponding electro-annealing durations are approximately 67 s, 82 s, and 100 s (Figure 1). The size of SEM images is 1 µm x1 µm.

Effect of Trace and Electrode Geometry on Temperature Distribution

In order to decouple the effect of varying constriction geometry from the non-uniform temperature

distribution, we employed an alternative parallel trace configuration identical to that used in Figure 4a also

shown in Figure S-5a below, except each trace has three identical constrictions (Figure S -5b). When this trace

network was electro-annealed, the coarsening was localized at the constrictions and as the color change

indicates, the influence of temperature distribution (observed in Figures 1d), was superimposed onto the

localized coarsening. More specifically, the constraints at the center were the most coarsened and the

coarsening at the constrictions decreased radially outward from the center.

Figure S-5. (a) Five parallel 1 mm-wide traces with progressively narrow constrictions from 0.9 mm (top) to 0.5 mm (bottom). (b) Five parallel 1 mm-wide traces with three 0.5 mm wide constrictions along the length of the trace.

References:

[1] Kurtulus, O.;Daggumati, P.; Seker, E. Molecular Release from Patterned Nanoporous Gold Thin Films. Nanoscale 2014, 6, 7062 - 7071.

[2] Dong, H.; Cao, X. Nanoporous Gold Thin Film: Fabrication, Structure Evolution, and Electrocatalytic Activity. The Journal of Physical Chemistry C 2008, 113, 603-609


mailto:[email protected]